How many legs does a horse have if you call a tail a
leg? Four. Calling a tail a leg does not make it a leg.
Calling something fire risk assessment
does not make it so either. Still, not every fire risk question requires
all the information you can muster. If the cost of more information
exceeds its value, a prudent engineer will use less.
My work has tended to emphasize more
sophisticated methods and the dangers involved if you settle for less.
My frequent partner, Jack Watts, has emphasized less sophisticated
methods and the value they can provide. We’re both right and both wrong,
on occasion. I’m going to try to channel both of us and provide an
overview of the spectrum of fire risk analysis methods.
A fire risk question starts with three
questions: What could happen? How bad could it be? How likely is it? A
fire risk decision adds a fourth: What should I do about that?
Checklists (including narratives)
Checklists comprehensively address what could happen (hazards) and what should be done about it, jumping over the other questions. This approach is required by law in the United Kingdom and used in "community fire risk assessment.” Checklists are the least expensive option. It is not surprising they are most used when fire risk assessment is routine.
Checklists can be useful prods to make plans
complete. Add a bit of information on how bad it can be, and you can
set inspection priorities. Think how useful this simple tool would have
been in West, Texas.
You can convert checklists into logic trees, showing the
interaction of different hazards and safeguards. A fault tree may show
there are alternative paths to safety.
Indexes are the first step into quantification. An index identifies elements that contribute to fire risk (hazards, safeguards) and rates the element (e.g., hazard severity or likelihood, safeguard effectiveness) and its importance (weights), often in a framework that shows system interactions. Quantification is subjective but systematic. An index rarely uses data but should be consistent with data. Some indexes have been withdrawn because they failed that test.
Indexes are used in the insurance
industry, dating back to the Dean schedule in 1902, in part because
there is no need to persuade multiple interests of the index’s accuracy
The index philosophy of simple displays of mostly subjective
estimates, systematically developed, also lives in FRA methods like risk
matrixes and risk curves.
Sophisticated fire risk assessment is embodied in the SFPE1, ISO2 and ASTM3 guides to fire risk assessment. This kind of complexity sends local authorities diving for copies of NFPA 5514, the guide to evaluating FRAs.
There are many reasons to favor
quantitative FRA. Likelihood and severity are examined explicitly.
Different kinds of fires are examined. The paper trail allows different
interests to critique and modify the analysis. There is enough
information for a conversation about tolerable risk. And the measures of
risk can be compared to costs.
There are many reasons to be uncomfortable
with quantitative FRA. Data requirements are enormous. Subjective
judgments are still needed in abundance. The physics must be simplified
to lower the calculation burden, and the physics is not self-sufficient,
because people and chance (ignition probabilities, reliabilities,
uncertainties, and human behavior) make too big a difference to ignore.
If you are going to comply with the code
anyway, the risks at stake aren’t big enough to justify the cost. If you
want to prove code equivalency, however, large risks are at stake and
you need a technical case strong enough to support a proposal – and
survive an NFPA 551 review.
NFPA’s performance-based initiative stresses that
use of sophisticated equivalency methods would be rare and exceptional.
Quantitative FRA should be even more rare and exceptional, but when you
need it, you really need it. No one should be expected to abandon
insistence on strict code compliance based on a few thin arguments
labeled as fire risk assessment.
John R. Hall, Jr. is with the National Fire Protection Association.
- SFPE Engineering Guide – Fire Risk Assessment, Society of Fire Protection Engineers, Bethesda, MD, November 2006.
- ISO 16732-1, Fire Safety Engineering—Guidance on Fire Risk Assessment, International Organization for Standardization, Geneva, Switzerland, 2012.
- ASTM E1776, Standard Guide for Development of Fire-Risk-Assessment Standards, ASTM International, West Conshohocken, PA, 2013.
- NFPA 551, Guide for the Evaluation of Fire Risk Assessments, National Fire Protection Association, Quincy, MA, 2013.